ADL: docking results not very accurate

Wilfred Li wilfred at
Thu Jan 31 10:33:17 PST 2008

I'd consider it reasonable if your 2nd best binding energy contains your
best pose. Would you have been able to select this cluster using some
other criteria, such as most populated cluster? Changing the rmsdtol may
also help. There are other settings to tweak, but you seem getting
close. There are no attachments, btw.



-----Original Message-----
From: autodock-bounces at [mailto:autodock-bounces at]
On Behalf Of Fabian Glaser
Sent: Thursday, January 31, 2008 12:40 AM
To: autodock at
Subject: ADL: docking results not very accurate

Dear Autodock colleagues,

I'd be very grateful for any advice with the following:

I am trying to perform docking experiments for a colleague. Initially I
have performed several experiments trying to re-dock the co-crystallized
ligand (the structure is 1nex, and the ligand, a modified peptide), to
check the procedure, and later the idea is to use the same procedure to
dock a peptide that they think represents the substrate on a modelled

The problem is that when I try to dock the co-crystallized ligand (as an
initial test), the "correct" result does not appear in the best energy
cluster: Attached figure snapshot3.png shows the best energetic cluster
found within 100 docked results, and snapshot2.png the second best
cluster, which is the correct  one (in blue the protein, in green the
co-crystallized ligand, the docked ligand is colored by element). The
different clusters are shown in the small figure on the right, in blue

I am quite a rocky with docking, so any help how can I improve these
results will be highly appreciated.

Best regards and thanks a lot in advance.


Fabian Glaser, PhD
Bioinformatics Knowledge Unit,
The Lorry I. Lokey Interdisciplinary
Center for Life Sciences and Engineering Technion - Israel Institute of
Technology Haifa 32000, ISRAEL

E-mail: fglaser at
Tel:   +972-(0)4-8293701
Cel:   +972-(0)54-4772396


More information about the autodock mailing list